The Negation Of The Criminal Responsibility Of The Artificial Intelligent Agents

Since the appearance of artificial intelligent agents in the 1950 s,as a result of artificial intelligence technology,artificial intelligent agents have passed through lots of changes with the ups and downs of artificial intelligence technology.Now using the top-down learning methods,the artificial intelligent agents have established a high degree of autonomy.However,this type of artificial intelligent agents has brought human not only more convenient and sufficient lifestyle,but also difficulties of the attribution of liability for damage,such as the accidental injury caused by surgical robot or the death caused by the autonomous car.In order to solve this problem,some scholars classify artificial intelligent agents into three categories: weak artificial intelligent agents,the smarter intelligent agents and the super intelligent agents,they believe that the smarter intelligent agents could have free will and against human-beings without the control of the human,so the smarter intelligent agents should take criminal liability by themselves,but the negative view criticize the affirmative view,the reason is that the affirmative view challenges the existing criminal law system which based on the anthropocentrism,and it is impossible for the smarter intelligent agents to appear.Based on the Negative viewpoint,this paper is divided into four parts.The first part of this paper discusses the concept,classification and the difficulty caused by the artificial intelligent agents.After sorting theories,there is a problem that some scholars do not take a clear definition of the artificial intelligent agents in their article or the classification of the artificial intelligent agents is not universal but only to support their arguments.Therefore,the first part of this paper is the concept and classification of the artificial intelligent agents.The object of this paper is the top-down artificial intelligent agents based on artificial neural network and deep learning algorithm,which is actually the smarter intelligent agents classified by the affirmative.The second and the third part based on the presentation and analysis of the two theories and the viewpoint of this paper.The determination of criminal responsibility has two conditions,one is the subjective dimension,another is the capacity,the former belongs to constitutive elements of crime,and the latter belongs to responsibility.Therefore,this paper negates the criminal liability of artificial intelligent agents from the perspective of subjectivity and the capacity.In the second part,the author sorts out the theories of the subjectivity of artificial intelligent agents in detail,and negates the subjectivity of artificial intelligent agents respectively from the perspective of natural person and the perspective of the legal person.In the third part,the author sorts out the theories of the criminal responsibility capacity of the artificial intelligent agents and negates the criminal responsibility capacity of the artificial intelligent agents from the perspective of cognitive and volitional capacity.The fourth part is the analysis of who should be responsible for the harmful result caused by the autonomous behavior of the artificial intelligent agents from the perspective of the negative view.